HYBRID PARTICLE SWARM OPTIMIZATION DAN K-MEANS CLUSTERING UNTUK PEMETAAN MUTU PENDIDIKAN BERDASARKAN ACUAN STANDAR NASIONAL PENDIDIKAN

Abstract

Lembaga pendidikan di Indonesia selama ini hanya menggunakan nilai akreditasi sekolah sebagai tolok ukur untuk mengetahui seberapa baik capaian mutu sekolah. Sekolah juga hanya mendapatkan nilai akhir dari proses akreditasi sekolah yang dilakukan oleh Lembaga Badan Akreditasi Nasional tanpa adanya evaluasi tingkat lanjut. Sekolah memerlukan rekomendasi perbaikan sebagai bentuk evaluasi sehingga mutu sekolah dapat mengalami peningkatan. Evaluasi mutu sekolah memerlukan standar acuan penilaian berupa standar nasional Pendidikan. Penerapan standar pendidikan dalam pemetaan mutu pada tingkat sekolah menengah pertama dapat menjadi informasi yang berguna untuk mengevaluasi seberapa jauh capaian mutu pendidikan pada setiap sekolah. Penelitian ini menggunakan metode Hybrid particle swarm optimization dan k-Means clustering untuk mengelompokan data standar pendidikan ke dalam 3 kategori yaitu; standar pendidikan Kategori satu, Standar pendidikan Kategori dua dan standar pendidikan Kategori tiga. Keluaran dari sistem akan menampilkan hasil pengelompokan capaian mutu ke dalam tiga kelompok kategori Standar pendidikan. Setiap sekolah akan diberikan rekomendasi berdasarkan standar Lembaga pendidikan sebagai bentuk evaluasi diri. Penelitian ini menggunakan dua belas skenario uji coba parameter PSO dengan inertia 0.4 - 0.9 dan social dan cognitive 0.14 dan 1.14. Hasil pengelompokan menunjukkan bahwa pada kriteria pertama terdapat 5 sekolah berada di klaster satu, 38 sekolah berada di klaster dua dan 106 sekolah berada di klaster tiga. Pada kriteria kedua terdapat 12 sekolah berada di klaster satu, 72 sekolah berada di klaster dua dan 65 sekolah berada di klaster tiga. Pada kriteria ketiga terdapat 5 sekolah berada di klaster satu, 45 sekolah berada di klaster dua dan 99 sekolah berada di klaster tiga. Pada kriteria keempat terdapat 4 sekolah berada di klaster satu, 56 sekolah berada di klaster dua dan 89 sekolah berada di klaster tiga. Pada kriteria kelima terdapat 24 sekolah berada di klaster satu, 67 sekolah berada di klaster dua dan 58 sekolah berada di klaster tiga. Kata kunci : Particle Swarm optimization, k-Means clustering, data Mining, pemetaan mutu Pendidikan. Educational institutions in Indonesia have only used the value of school accreditation as a benchmark to find out how well school quality is achieved. Schools also only get the final grade from the school accreditation process carried out by the National Accreditation Agency without any further evaluation. Schools need improvement recommendations as a form of evaluation so that school quality can increase. Evaluation of school quality requires an assessment standard of reference in the form of national education standards. The implementation of education standards in mapping quality at the junior high school Kategori can be useful information to evaluate how far the quality of education is achieved at each school. This study uses Hybrid particle swarm optimization and k-means clustering methods to classify education standard data into 3 categories, namely; category one education standards, category two education standards and category three education standards. The output of the system will display the results of grouping quality achievements into three categories of education standard categories. Each school will be given a recommendation based on the standard of an educational institution as a form of self-evaluation. This study uses twelve scenarios of PSO parameters with inertia 0.4 - 0.9 and social and cognitive 0.14 and 1.14. grouping results show that in the first criteria there were 5 schools in cluster one, 38 schools were in the second cluster and 106 schools were in the third cluster. In the second criterion there were 12 schools in cluster one, 72 schools were in the second cluster and 65 schools were in the third cluster. In the third criterion there were 5 schools in cluster one, 45 schools were in the second cluster and 99 schools were in the third cluster. In the fourth criterion there are 4 schools in cluster one, 56 schools are in the second cluster and 89 schools are in the third cluster. In the fifth criterion there were 24 schools in cluster one, 67 schools were in the second cluster and 58 schools were in the third cluster. Keywords: Particle Swarm optimization, k-Means clustering, data Mining, mapping the quality of Education.